Concepts
An entity bean models a business entity and performs actions
within a business process. For example, you could implement an entity bean
to retrieve and perform computation on line items within a purchase order.
Entity beans survive beyond the end of an application session, even a server
crash or a network failure, because the data is stored by the container in
some form of data storage system, such as a database.
The EJB specification defines the following persistence
mechanisms for entity beans:
-
Bean-manged persistence (BMP), where data is
stored and retrieved programmatically through methods implemented implemented
by the bean developer. These methods typically use JDBC or SQLJ to manage
persistence.
- Container-managed persistence (CMP), where the container
handles persistence chores automatically.
Container-managed persistence keeps an entity bean class separate
from its persistent representation, enabling developers to change a bean's data
source without affecting its implementation. With CMP, instead of writing Java
code to implement bean-managed persistence, you describe container-managed persistence
declaratively in the deployment descriptor file (for an example from the FBS 10g,
see ibfbs\etc\ejb-jar.xml). When a bean is deployed, the container
provider's tools parse the deployment descriptor and generate code to implement
the underlying classes. See the Design and Implementation
sections of this tutorial for more information and examples.
At runtime, the container manages the bean's data by interacting
with the datasource, typically a relational database, designated in the deployment
descriptor. If the persistent data is saved to or restored from a database,
you must ensure that the correct tables exist for the bean—if you use OC4J,
you can configure the container to create such tables automatically.
|
Bookmark
